Warmer temperatures 1000 and 2000 years ago

[Little Fish]

I was curious why you said peter wegman.

did you seriously get your information here:

http://answers.yahoo.com/question/index?qid=20111005092637AA0Wtnn

[questionmark]

I was curious why you said peter wegman.

did you seriously get your information here:

http://answers.yahoo...05092637AA0Wtnn

Because occasionally I commit an error too, I was thinking about Peter Christoff but then remembered that he hardly is s scientist but a journalist.

[spud the mackem]

Just curious, where did these graphs originate and how were they made up, and by what resources.Anyone can stand in front of a rolling graph chart and cause the stylus to shake then come up with a theory.So where did the evidence come from ??

[Zeta Reticulum]

Earth has been cooling and warming for millions and millions of years..... ICE AGE ... thaw... ICE AGE... thaw.... now is nothing different

[questionmark]

Just curious, where did these graphs originate and how were they made up, and by what resources.Anyone can stand in front of a rolling graph chart and cause the stylus to shake then come up with a theory.So where did the evidence come from ??

There are different types of reconstructions, based on tree growth, sediment analysis and ice cores. We know the warmth induced biological and chemical processes well enough to give us an approximate picture of the temperatures that must have reigned at a specific time.

The easiest is to make a list of plants that thrived during a period, like the discovery of the Medieval Optimum, that was based mostly on the discovery that in Britain plants that needed much warmer temperatures than found at this time thrived.

More exact are the measurements of decomposition products found in lake sediments and most exact are gas analysis of the polar ice cores.

[spud the mackem]

There are different types of reconstructions, based on tree growth, sediment analysis and ice cores. We know the warmth induced biological and chemical processes well enough to give us an approximate picture of the temperatures that must have reigned at a specific time.

The easiest is to make a list of plants that thrived during a period, like the discovery of the Medieval Optimum, that was based mostly on the discovery that in Britain plants that needed much warmer temperatures than found at this time thrived.

More exact are the measurements of decomposition products found in lake sediments and most exact are gas analysis of the polar ice cores.

Ok but where did the graph come from ? It could be a siesmic graph,or a voice graph,or any thing you want it to be.

[questionmark]

Ok but where did the graph come from ? It could be a siesmic graph,or a voice graph,or any thing you want it to be.

That can be found in the paper linked to above:

The HadCRUT3v was downloaded from

http://www.cru.uea.ac.uk/cru/data/temperature/.

[Doug1029]

not true.

While Energy and Environment is a technical journal (technically), it has a pretty poor record when it comes to environment. I will post this example of an article by McKitrick and Michaels to illustrate the point:

In 2004, Ross McKitrick, an economist with close ties to fossil fuel interests, published a paper:

McKitrick, R. and P. J. Michaels. 2004. A test of corrections for extraneous signals in gridded surface temperature data. Climate Research 26(2004) 159-173.

The paper, published in Energy and Environment, they claimed, was the “death knell of human caused climate change.” Michaels characterized the paper as “a bombshell” destined to “knock the stuffing out of” the mainstream view that human activity caused observed surface warming. McKitrick and Michaels purported to demonstrate that contaminated temperature records based on a correlation they claimed to have found (but never demonstrated) between economic activity and temperature change in various regions of the globe.

Michaels claimed the paper was subject to “four years of the most rigorous peer reviews ever.” Weeks after publication, Tim Lambert, an Australian computer science expert, discovered a fatal flaw. In attempting to account for the effect of latitude, McKitrick had confused degrees and radians and had, in effect, been feeding his process random noise.

I won’t call this a grade-school mistake because spherical geometry is usually taught at the junior level in college math programs. McKitrick is an economist and spherical geometry is an alien world to him. By itself, this is enough to invalidate the study, but there were other mistakes as well.

This is only one of many dubious papers produced by McKtrick et al. and published in Energy and Enviroment. I propose to present a review of them, posting brief summaries and references here. As the list of articles refuting McKitrick is a lengthy one, I am wondering if I should start a new thread. Karlis: Opinions?

Doug

Edited October 22, 2012 by Doug1029

[Doug1029]

The paper, published in Energy and Environment, they claimed, was the “death knell of human caused climate change.” Michaels characterized the paper as “a bombshell” destined to “knock the stuffing out of” the mainstream view that human activity caused observed surface warming. McKitrick and Michaels purported to demonstrate that contaminated temperature records based on a correlation they claimed to have found (but never demonstrated) between economic activity and temperature change in various regions of the globe.

The paper above was published in Climate Research, not Energy and Environment. Oops!

Doug

[Little Fish]

michael mann's hockey stick graphs used by the IPCC and still held in the minds of zealous alarmists have been shown to be bad science and not representative of past temperatures, it's that simple. doug, going after the people that discovered and SHOWED this in published research on unrelated matters and going after the publisher that published it is just a political smear campaign. obviously when their research cannot be discredited, they attempt to discredit the messengers, its called denial. the greek philosophers sorted all that out thousands of years ago.

[Doug1029]

michael mann's hockey stick graphs used by the IPCC and still held in the minds of zealous alarmists have been shown to be bad science and not representative of past temperatures, it's that simple. doug, going after the people that discovered and SHOWED this in published research on unrelated matters and going after the publisher that published it is just a political smear campaign. obviously when their research cannot be discredited, they attempt to discredit the messengers, its called denial. the greek philosophers sorted all that out thousands of years ago.

There is a now sizeable body of published research vindicating Mann. That's what I am proposing to post. But as I said above, it's a lengthy one and will take up a lot of space.

If Mann's work has been seriously discredited, I have yet to see the article explaining that. You certainly haven't posted a reference to it.

If McKitrick's work is seriously flawed, how can pointing that out be a "smear campaign?" How can confusing degrees and radians not be a mistake?

Doug

[Doug1029]

michael mann's hockey stick graphs used by the IPCC and still held in the minds of zealous alarmists have been shown to be bad science and not representative of past temperatures, it's that simple. doug, going after the people that discovered and SHOWED this in published research on unrelated matters and going after the publisher that published it is just a political smear campaign. obviously when their research cannot be discredited, they attempt to discredit the messengers, its called denial. the greek philosophers sorted all that out thousands of years ago.

There is a now sizeable body of published research vindicating Mann. That's what I am proposing to post. But as I said above, it's a lengthy one and will take up a lot of space.

If Mann's work has been seriously discredited, I have yet to see the article explaining that. You certainly haven't posted a reference to it. That paper by McIntyre and McKitrick didn't do it.

In charging someone with flawed research, you have to get pretty specific: you have to say exactly what the mistake was and where so that future researchers don't repeat it. McIntyre and McKitrick's paper didn't do that. It claimed to, but... Here's an example from page 753:

"(a) unjustified trancation of 3 series;"

A series is just the list of ring widths produced in the reading of one tree's rings. M&M do not say what Mann's justification was, nor do they give one of their own. In their unsubstantiated OPINION this was a mistake, but they do not support their claim.

I have truncated series - regretably, I had to truncate my oldest series. The last four rings were pitch loaded and I could not be sure that they weren't false rings. Also, it takes a minimum number of series, at least thirteen, to get a strong enough expressed population signal (EPS must be greater than 0.85.) to be acceptable for climate work. The farther back you go in any chronology, the fewer series there are with which to estimate average ring thickness. Thirteen is the minimum standard, but it usually takes more to get the 0.85 EPS. I have several chronologies with over a hundred years of data that can't be used because I can't guarantee their accuracy due to the small sample size. When part of a series does not intercorrelate with the other series (35% minimum correlation), it is truncated. The good part is used while the rest is discarded.

There are legitinate reasons to truncate a series. If M&M are going to say that truncation was not appropriate, they need to say why.

"( copying 1980 values from one series onto other series, resulting in incorrect values in at least 13 series;"

"Three are PCs from Texas-Oklahoma sites." I work with those very chronologies and will be adding eleven new chronologies to the literature within a year. The paper is already written, but not yet back from peer review. PM me with your email address for a pre-publication copy.

1980 in Oklahoma, Arkansas and Texas was a severe drought year. Series from rocky sites are often missing the 1980 ring. The drought broke on September 28 with a deluge. Trees were not quite dormant and responded with a narrow ring of latewood cells with an absolutely unique appearance: very regular lumens and the ring only two cells in thickness. But the rain came too late for many trees and they did not produce a ring at all that year.

One does not just copy values from one series to another. If the ring has a zero thickness, you put zero in the blank. If you then detrend the series, the result is that zero values are replaced by negative values. Some dendrochronologists don't like this and replace the negatives with zeroes. I have no problem with negative rings: all detrended values are hypothetical anyway, so what's the problem with a negative one? You then run the series through a tranform process in which you multiply each ring width by the grand mean for the chronology and then divide it by the average ring width of the series. This gives each series equal weight, so that you aren't giving a fast-growing tree more weight than a slower-growing tree. You then average the ring widths by year to produce the chronology. The averaging process eliminates the negatives - usually (I've never seen an exception, but it is theoretically possible.). Apparently, M&M do not like replacement of negative values with zeroes prior to weight equalization (I don't either, but there's nothing technically wrong with it.). Also, we're talking about a dataset reaching back to 1400. The error caused by replacing thirteen negatives with zeroes will not even be detectable in the resulting temperature estaimates, particularly since those estimates were ten-year running averages.

M&M apparently missed the little detail that 1866, 1871, 1879, 1881, 1886, 1894, 1909, 1916, 1918, 1930, 1936, 1943, 1956,1963, 1978 and 1992/1993 had severe winter storms that resulted in many trees with missing rings for those years. The missing rings would have produced negative values that would have been replaced by zeroes.

"© displacement of 18 series to one year earlier than apparently intended;"

That's easy to do. I have many times examined a series and noted sections of it that seem to be a year off from the master chronology. But as I scan down the series, it comes back into sync. That could be caused by a missing ring followed several years later by a false ring. The result would shift a section of the series to a position one year earilier than it actually was. That's a cross-dating error and would have been present in the original chronology. M&M could easily have checked this, but didn't.

OR: the apparent error could have been caused by the very patchy nature of winter storms in the central US. A storm that smashes forests in one place may not even break a twig 30 miles away. I lost 48 trees on 0.4 acres at one site and two on a similar site five miles away. The patchy nature of winter storms often produces ring sequences that are one ring off the regional chronology. To find if this is true, compare chronologies from different areas in the same region. The shift will be consistent within a chronology but not between chronologies. M&M could have checked for this, but didn't. My take: this is probably not a mistake at all; I could double check this if M&M had told me which of those 18 series they thought had the problem. But they didn't.

"(d)" unjustified extrapolations or interpolations to cover missing entries in 19 series;"

M&M don't say what they think would be justified. I don't think ANY such extrapolation is justified if you have a large sample size and Mann's paper used thousands of series - it was more than big enough. But many other dendrochronologists don't believe this, so much so, that Cook's computer program, ARSTAN, automatically fills missing data. How did this get into Mann's work? Mann used chronologies prepared by other people; he did not prepare them himself. It was easier than recalculating all those chronologies again. It was also standard practice, of a sort, even if its use is not universally accepted in dendrochronology. It is not "unethical" or "improper;" this is why Mann can be exhonerated without having used the best of techniques.

The important issue is whether this invalidates the hockey stick. Because the missing values are calculated using real measurements from other series, the estimated thickness of the missing ring is probably very close to the actual thickness. Once again, the error introduced by the method is tiny and would need to be thousands of times larger to produce a measuremable change in the slope of the temperature line.

"(e) geographical mislocations and missing identifiers of location."

Geographical location is only used if data is being corrected for unequal spatial distributions, or if one is looking for spatial autocorrelation. Mann was doing neither. Neither data nor results were gridded. Other than these two instances, geographical information is not used and is irrelevant. M&M are, in effect, asking Mann why he didn't list the number of cups of coffee he drank each morning - it has nothing to do with the issue!

"(f) inconsistent use of seasonal data where annual data are available;"

This is a valid question. It is not a concern when data for the same season are used throughout a series, but if seasonal data is used for part of a series, but annual data is used for another part, then there is a discontinuity at the point where the two join. This can create false temperature discrepancies between the two systems. M&M's description at the bottom of the page suggests that data were used consistently within a series, thus, admitting the use was legitimate.

"(g) obsolete data in at least 24 series, some of which may have been already obsolete at the time of the MBH98 calculations;"

I have to admit it: I don't understand this one. How does a series become obsolete? The measurements don't change while the core is sitting on the shelf. The only thing that could change is the standard method for processing the data. But that's easily fixed by running the old data through the new process. I'm at a loss.

"(h) listing of unused proxies;"

When this is done, there's usually a disclaimer attached. Journals don't like to dedicate space to this sort of thing, so authors don't usually do it. Sounds like somebody misread the Mann paper, but as I don't have a copy at the moment, I can only speculate.

"(i) incorrect calculation of all 28 principal components."

I am not qualified to comment on PCA. And, I have not had a chance to sort through papers by people who do. It is my understanding that this charge was specifically refuted. I'll need to do some more studying here.

I said this would be a lengthy topic. And this is only one page of one paper.

I have pointed out some flaws in McKitrick's charges against Mann. How can simply pointing out technical mistakes constitute a "smear campaign?" How can confusing degrees and radians not be a mistake?

The ethical thing for McKitrick to do when the mistake was brought to his attention was to voluntarily withdraw his paper. This he has not done. Neither has he offered any defense of his paper. But mistakes happen. By itself, a mistake is not necessarily unethical.

Doug

Edited October 24, 2012 by Doug1029

[Guest]

How Little Fish can defend the McIntyre and McKitrick paper when its multiple gross flaws have many times been discussed at length on here - and use it as his sole basis for questioning the Mann Hockey Stick, shows that he is incapable of assimilating information and facing up to the fact that his position is based on a pyramid of falsifications.

As such Doug - I genuinely feel you are wasting your valuable time.

Br Cornelius

Edited October 24, 2012 by Guest

[Little Fish]

"(a) unjustified trancation of 3 series;"

A series is just the list of ring widths produced in the reading of one tree's rings. M&M do not say what Mann's justification was, nor do they give one of their own. In their unsubstantiated OPINION this was a mistake, but they do not support their claim.

I have truncated series - regretably, I had to truncate my oldest series. The last four rings were pitch loaded and I could not be sure that they weren't false rings. Also, it takes a minimum number of series, at least thirteen, to get a strong enough expressed population signal (EPS must be greater than 0.85.) to be acceptable for climate work. The farther back you go in any chronology, the fewer series there are with which to estimate average ring thickness. Thirteen is the minimum standard, but it usually takes more to get the 0.85 EPS. I have several chronologies with over a hundred years of data that can't be used because I can't guarantee their accuracy due to the small sample size. When part of a series does not intercorrelate with the other series (35% minimum correlation), it is truncated. The good part is used while the rest is discarded.

There are legitinate reasons to truncate a series. If M&M are going to say that truncation was not appropriate, they need to say why.

the truncated series referred to are instrumental series, not tree rings chronologies.

how did you make this mistake? LOL.

hey look, cornelius "likes your post", see what he does - he supports an argument when he doesn't even understand it.

"b. copying 1980 values from one series onto other series, resulting in incorrect values in at least 13 series;"

"Three are PCs from Texas-Oklahoma sites." I work with those very chronologies and will be adding eleven new chronologies to the literature within a year. The paper is already written, but not yet back from peer review. PM me with your email address for a pre-publication copy.

1980 in Oklahoma, Arkansas and Texas was a severe drought year. Series from rocky sites are often missing the 1980 ring. The drought broke on September 28 with a deluge. Trees were not quite dormant and responded with a narrow ring of latewood cells with an absolutely unique appearance: very regular lumens and the ring only two cells in thickness. But the rain came too late for many trees and they did not produce a ring at all that year.

One does not just copy values from one series to another. If the ring has a zero thickness, you put zero in the blank. If you then detrend the series, the result is that zero values are replaced by negative values. Some dendrochronologists don't like this and replace the negatives with zeroes. I have no problem with negative rings: all detrended values are hypothetical anyway, so what's the problem with a negative one? You then run the series through a tranform process in which you multiply each ring width by the grand mean for the chronology and then divide it by the average ring width of the series. This gives each series equal weight, so that you aren't giving a fast-growing tree more weight than a slower-growing tree. You then average the ring widths by year to produce the chronology. The averaging process eliminates the negatives - usually (I've never seen an exception, but it is theoretically possible.). Apparently, M&M do not like replacement of negative values with zeroes prior to weight equalization (I don't either, but there's nothing technically wrong with it.). Also, we're talking about a dataset reaching back to 1400. The error caused by replacing thirteen negatives with zeroes will not even be detectable in the resulting temperature estaimates, particularly since those estimates were ten-year running averages.

M&M apparently missed the little detail that 1866, 1871, 1879, 1881, 1886, 1894, 1909, 1916, 1918, 1930, 1936, 1943, 1956,1963, 1978 and 1992/1993 had severe winter storms that resulted in many trees with missing rings for those years. The missing rings would have produced negative values that would have been replaced by zeroes.

again, these are not tree ring chronologies as you are claiming. These series refer to computed PCs. are you and cornelius suggesting it is ok to make up the results of calculations??

In the MBH98 collated data set the 1980 values for series #72-#80, which are

the 9 Texas-Mexico principal components computed by MBH98, are identical to 7

decimal places, an obviously impossible result (see Table 1)

http://www.multi-sci...e-mckitrick.pdf

and you say Mann's paper used thousands of series - Mann's paper uses 112 series, 28 of which are computed PCs, I would conclude you have not read Mcintyre's paper which you are commenting on.

"c) displacement of 18 series to one year earlier than apparently intended;"

That's easy to do. I have many times examined a series and noted sections of it that seem to be a year off from the master chronology. But as I scan down the series, it comes back into sync. That could be caused by a missing ring followed several years later by a false ring. The result would shift a section of the series to a position one year earilier than it actually was. That's a cross-dating error and would have been present in the original chronology. M&M could easily have checked this, but didn't.

OR: the apparent error could have been caused by the very patchy nature of winter storms in the central US. A storm that smashes forests in one place may not even break a twig 30 miles away. I lost 48 trees on 0.4 acres at one site and two on a similar site five miles away. The patchy nature of winter storms often produces ring sequences that are one ring off the regional chronology. To find if this is true, compare chronologies from different areas in the same region. The shift will be consistent within a chronology but not between chronologies. M&M could have checked for this, but didn't. My take: this is probably not a mistake at all; I could double check this if M&M had told me which of those 18 series they thought had the problem. But they didn't.

again not tree ring chronologies, but computed principal component series.

"(d)" unjustified extrapolations or interpolations to cover missing entries in 19 series;"

M&M don't say what they think would be justified. I don't think ANY such extrapolation is justified if you have a large sample size and Mann's paper used thousands of series - it was more than big enough. But many other dendrochronologists don't believe this, so much so, that Cook's computer program, ARSTAN, automatically fills missing data. How did this get into Mann's work? Mann used chronologies prepared by other people; he did not prepare them himself. It was easier than recalculating all those chronologies again. It was also standard practice, of a sort, even if its use is not universally accepted in dendrochronology. It is not "unethical" or "improper;" this is why Mann can be exhonerated without having used the best of techniques.

The important issue is whether this invalidates the hockey stick. Because the missing values are calculated using real measurements from other series, the estimated thickness of the missing ring is probably very close to the actual thickness. Once again, the error introduced by the method is tiny and would need to be thousands of times larger to produce a measuremable change in the slope of the temperature line.

you assume that data was filled in where data was not available, not true.

"MBH98 insert extrapolated, interpolated or copied values during the critical

calibration period into 19 series. We refer to these as “fills” hereafter. In the data set

provided to the authors, the following 17 series contain end-of-sample fills for one or

more years including 1980: #6, #45, #46, #50-#52, #54-#56, #58, #93-#99. Series #53

was filled for 4 years at its beginning and series #3 for 16 years in the calibration

period. In the case of #3, MBH98 inexplicably replaced available source values for

1962-64 with filled values. For examples see Tables 3, 4 and 5.

Series #50 is especially noteworthy. The values of series #50 for the entire period

from 1962 to 1982 are copied from series #49 (see Table 6). Although MBH98

attribute both series #49 and #50 to Fritts and Shao (1992), series #49 is actually

derived from Briffa et al. (1992)."

http://www.multi-sci...e-mckitrick.pdf

"(e) geographical mislocations and missing identifiers of location."

Geographical location is only used if data is being corrected for unequal spatial distributions, or if one is looking for spatial autocorrelation. Mann was doing neither. Neither data nor results were gridded. Other than these two instances, geographical information is not used and is irrelevant. M&M are, in effect, asking Mann why he didn't list the number of cups of coffee he drank each morning - it has nothing to do with the issue!

it might seem irrelevant if you are arguing against your personal assumptions, but you are paying no mind to the peer review process. the issue here is that some data used by the Mann paper cannot be identified.

"MBH98 use 11 precipitation series, for which they cite Jones and

Bradley (1992) (hereafter “JB92”). JB92 (Table 13.3) lists 17 precipitation series, of

which 12 are digitally published at the World Data Center for Paleoclimatology

(http://www.ngdc.noaa...aleo/paleo.html, hereinafter denoted WDCP). In only

two MBH98 precipitation series (#35 and #37) did the correlation between JB92 and

MBH98 data exceed 0.9, permitting a reasonably secure identification of locations;

other correlations were less than 0.5 excluding the possibility of identification. The

JB92 series for Paris, France (48.8N, 2.5E) can be identified with MBH98 series #37

both from the high correlation and the identity of starting date (see Figure 2, which

graphs both these series). However, MBH98 series #37 is located at the grid-box

centred at 42.5N, 72.5W near Boston, Massachusetts.

Two MBH98 precipitation series are in India and derive from an unreported source,

since no Indian locations are listed in JB92. The other 7 MBH98 precipitation series

derive either from unreported sources, from the 5 JB92 series not digitally published

at WDCP or have been heavily transformed in collation. Two of the MBH98

temperature grid-box series had no locational counterparts in JB92 (Table 13.1): series

#26 (52.5N, 17.5E grid-box) and series #29 (62.5N, 7.5E grid-box). In addition,

MBH98 series #20 (Central Greenland ice core) is materially mislocated to the north

and west. On comparison with source data, it can be seen that MBH98 have also

reversed the geographical locations of series #46 and #47."

if you read the above passage you have to wonder why this simple errors of identification were not picked up in peer review, you have to conclude it was not reviewed adequately. it is required for peer review to get this right, you cannot trivialize unidentified and misidentified data by your cups of coffee analogy. are you suggesting that after discovering these errors and thus knowing of them M&M should have ignored these errors?

"(f) inconsistent use of seasonal data where annual data are available;"

This is a valid question. It is not a concern when data for the same season are used throughout a series, but if seasonal data is used for part of a series, but annual data is used for another part, then there is a discontinuity at the point where the two join. This can create false temperature discrepancies between the two systems. M&M's description at the bottom of the page suggests that data were used consistently within a series, thus, admitting the use was legitimate.

you make the mistake of assuming that annual data was joined onto seasonal data to form a longer series. what occured was that Mann used only summer temperatures from central europe to calculate an annual series when annual data air temperature data for central europe was available. this is considered cherry picking. see Figure 1.

http://www.multi-sci...e-mckitrick.pdf

"(g) obsolete data in at least 24 series, some of which may have been already obsolete at the time of the MBH98 calculations;"

I have to admit it: I don't understand this one. How does a series become obsolete? The measurements don't change while the core is sitting on the shelf. The only thing that could change is the standard method for processing the data. But that's easily fixed by running the old data through the new process. I'm at a loss.

"datasets used by MBH98 were already obsolete in 1998. In response to an inquiry

about series #51- #61, WDCP confirmed that the updated versions for four of the

series were available as early as 1991-1992"

maybe you should contact the World Data Center for Paleoclimatology (WDCP) to find out why some of Mann's tree ring data was superseded seven years before his study.

"(h) listing of unused proxies;"

When this is done, there's usually a disclaimer attached. Journals don't like to dedicate space to this sort of thing, so authors don't usually do it. Sounds like somebody misread the Mann paper, but as I don't have a copy at the moment, I can only speculate.

there is no reason to list proxies in the study and then not use them in the study.

"when this is done" - bah! what Bothersome Stuff.

"Five series purportedly in the multiproxy network (fran003, ital015, ital015x, spai026 and spai047 in the MBH98 list “ITRDB –Miscellaneous”) cannot be located in either the MBH98 collated set or the proxy PC compilations."

"(i) incorrect calculation of all 28 principal components."

I am not qualified to comment on PCA. And, I have not had a chance to sort through papers by people who do. It is my understanding that this charge was specifically refuted. I'll need to do some more studying here.

handwaving since that is the most serious discovery and acknowledged by multiple independent statistics experts.

I have pointed out some flaws in McKitrick's charges against Mann.

no you haven't.

How can simply pointing out technical mistakes constitute a "smear campaign?" How can confusing degrees and radians not be a mistake?

whatever you think you found it relates to a different paper published in a different journal and by different coauthors. it is an ad hominem argument, going after the man and not the ball is smearing which is what you did because that issue has nothing to do with this paper under discussion.

The ethical thing for McKitrick to do when the mistake was brought to his attention was to voluntarily withdraw his paper. This he has not done. Neither has he offered any defense of his paper. But mistakes happen. By itself, a mistake is not necessarily unethical.

you have highlighted no mistake in this paper. I think you are trying to imply this "radians" thing relates to this M&M paper when it doesn't. given the mistakes and incorrect assumptions you made here I have to doubt the assertions you make regarding this separate issue. Edited October 25, 2012 by Little Fish

[Doug1029]

[/color]

the truncated series referred to are instrumental series, not tree rings chronologies.

how did you make this mistake? LOL.

According to Mann et al. (1998), the series were "multiproxy series" which included "annual resolution dendroclimatic, ice core, ice melt and long historical records combined with other coral, ice and instrumental records." "Dendroclimatic" records are derived from tree rings.

This is a direct quote from Mann's errata sheet to the original Supplementary Information. It was published in Nature:

"Series (2) used in ref. 1 but not listed in original Supplementary Information:

Unpublished Southwest US/Mexico Density series (D. W. Stahle, personal communication)

Unpublished Southwest US/Mexico Latewood Width series (D. W. Stahle, personal communication)"

David W. Stahle is a dendrochronologist at the University of Arkansas and was an ah hoc member of my Ph.D. committee. That Texas-Mexico series is based on TREE RING DATA.

again, these are not tree ring chronologies as you are claiming. These series refer to computed PCs.

Principal component analysis includes data from multiple proxies. Tree ring data, ice core data, instrumental data can all be used to create one PCA series.

It is becoming clear that you know even less about PCA than I do. It would help a whole lot if you would actually read Mann's paper.

are you and cornelius suggesting it is ok to make up the results of calculations??

Truncation involves the deletion of all data from a particular point either forward or backward to the end of the series. It does not involve making up data. It is a valid technique when the data being deleted has failed intercorrelation tests or for whatever reason is not reliable (like in the curled grain around a knot or fire scar).

In the MBH98 collated data set the 1980 values for series #72-#80, which are

the 9 Texas-Mexico principal components computed by MBH98,

The Texas-Mexico principal components were computed from David Stahle's tree ring chronologies.

are identical to 7

decimal places, an obviously impossible result (see Table 1)

http://www.multi-sci...e-mckitrick.pdf

When you replace negative values with zeroes, the results are identical to infinity. Seven places sounds like a computer-generated rounding error resulting from using a floating point value in an integer format. In most cases, you can igore this, particularly in climatology which isn't accurate to seven places, anyway.

and you say Mann's paper used thousands of series - Mann's paper uses 112 series, 28 of which are computed PCs, I would conclude you have not read Mcintyre's paper which you are commenting on.

According to Mann's paper, the multiproxy network consisted of 112 grid point series. Each of those grid point series is computed from multiple proxies. In the case of tree rings, these proxies were chronologies. Each chronology is computed from multiple tree ring series (minimum of 13). Mann used the Lake Winona (28 series), the Hots Springs (16 series) and McCurtain County (31 series), among many others. To find out exactly which ones those others were, google: NCDC tree rings. Click on the search engine and search for chronologies in Texas and Mexico by D. W. Stahle. Then count up the number of series in each. Bet it's in the hundreds, at least.

again not tree ring chronologies, but computed principal component series.

Again, you need to learn how PCAs are computed.

you assume that data was filled in where data was not available, not true.

"MBH98 insert extrapolated, interpolated or copied values during the critical

calibration period into 19 series. We refer to these as “fills” hereafter. In the data set

provided to the authors, the following 17 series contain end-of-sample fills for one or

more years including 1980: #6, #45, #46, #50-#52, #54-#56, #58, #93-#99. Series #53

was filled for 4 years at its beginning and series #3 for 16 years in the calibration

period. In the case of #3, MBH98 inexplicably replaced available source values for

1962-64 with filled values. For examples see Tables 3, 4 and 5.

Series #50 is especially noteworthy. The values of series #50 for the entire period

from 1962 to 1982 are copied from series #49 (see Table 6). Although MBH98

attribute both series #49 and #50 to Fritts and Shao (1992), series #49 is actually

derived from Briffa et al. (1992)."

http://www.multi-sci...e-mckitrick.pdf

To use filled data in a series that already has good data makes no sense at all. All you're doing is making a lot of extra work for yourself. An "end of sample fill" is simply filling missing data values with estimated ones. If the estimates are of high quality, there is no problem. If they aren't, they shouldn't be used at all. This would be a good reason to truncate a series, rather than accept a poor correlation.

it might seem irrelevant if you are arguing against your personal assumptions, but you are paying no mind to the peer review process. the issue here is that some data used by the Mann paper cannot be identified.

That was a mistake on my part. Mann used gridded series. That means he had to have location information, otherwise he could not have done the calculations (Try to find y when you don't know what x is.). As McKitrick pointed out, some of this was missing in Mann's paper. The Supplementary Information was not printed with the original article (to save space) but is available on Nature's website (www.nature.com/nature). The component series and their locations are listed there. Mann is quite clear that the gridded series were spaced at 5-degree intervals. The grid locations are at each multiple of 5 degrees of latitude and longitude. Mann did not say exactly which ones he used and which ones he left out, but with only 112 points, he couldn't have included them all. His proxies all come from temperate areas, so this is, presumably, where they are, a temperate-zone reconstruction (A polar one would have shown a greater slope to the handle of the "hockey stick.").

You bring up a good point about peer review. The reviewers should have caught that and sent the article back for revision. But peer reviwers are not paid (They do it for browny points because it looks good on a resume.). Publishers are complaining about a shortage of reviewers and often ask authors to suggest reviewers. Nuff said.

"MBH98 use 11 precipitation series, for which they cite Jones and

Bradley (1992) (hereafter “JB92”). JB92 (Table 13.3) lists 17 precipitation series, of

which 12 are digitally published at the World Data Center for Paleoclimatology

(http://www.ngdc.noaa...aleo/paleo.html, hereinafter denoted WDCP). In only

two MBH98 precipitation series (#35 and #37) did the correlation between JB92 and

MBH98 data exceed 0.9, permitting a reasonably secure identification of locations;

other correlations were less than 0.5 excluding the possibility of identification. The

JB92 series for Paris, France (48.8N, 2.5E) can be identified with MBH98 series #37

both from the high correlation and the identity of starting date (see Figure 2, which

graphs both these series). However, MBH98 series #37 is located at the grid-box

centred at 42.5N, 72.5W near Boston, Massachusetts.

Two MBH98 precipitation series are in India and derive from an unreported source,

since no Indian locations are listed in JB92. The other 7 MBH98 precipitation series

derive either from unreported sources, from the 5 JB92 series not digitally published

at WDCP or have been heavily transformed in collation. Two of the MBH98

temperature grid-box series had no locational counterparts in JB92 (Table 13.1): series

#26 (52.5N, 17.5E grid-box) and series #29 (62.5N, 7.5E grid-box). In addition,

MBH98 series #20 (Central Greenland ice core) is materially mislocated to the north

and west. On comparison with source data, it can be seen that MBH98 have also

reversed the geographical locations of series #46 and #47."

if you read the above passage you have to wonder why this simple errors of identification were not picked up in peer review, you have to conclude it was not reviewed adequately. it is required for peer review to get this right, you cannot trivialize unidentified and misidentified data by your cups of coffee analogy. are you suggesting that after discovering these errors and thus knowing of them M&M should have ignored these errors?

Agreed that the reviewers missed a lot of this. Mann et al. published a list of errors and provided much of this information, but it came out after M&M's article. It is available at the Nature site I referenced above. It is now eight years since the errata list was published. These objections have been met.

you make the mistake of assuming that annual data was joined onto seasonal data to form a longer series. what occured was that Mann used only summer temperatures from central europe to calculate an annual series when annual data air temperature data for central europe was available. this is considered cherry picking. see Figure 1.

http://www.multi-sci...e-mckitrick.pdf

What Mann did was to use some seasonal series along with annual series to create the PCs. You evidently missed the point I was trying to make about not using the two together within a single series. It's not cherry-picking, but it does weaken the correlations. One would not choose to do it if uniform data sets were available. Was Mann aware that there were better datasets? I don't know.

"datasets used by MBH98 were already obsolete in 1998. In response to an inquiry

about series #51- #61, WDCP confirmed that the updated versions for four of the

series were available as early as 1991-1992"

maybe you should contact the World Data Center for Paleoclimatology (WDCP) to find out why some of Mann's tree ring data was superseded seven years before his study.

I did better than that. I contacted Mann. He used the raw data. Most scientists won't use somebody else's chronology unless they can determine exactly how it was processed. Because the raw data is published along with the chronology, one can go back and reprocess it, in effect, producing an updated version. Like I said above, the data doesn't change while the core is sitting on the shelf.

there is no reason to list proxies in the study and then not use them in the study.

"when this is done" - bah! what Bothersome Stuff.

Maybe if you had read that paragraph, you'd understand that I was agreeing with you. This should have been caught by the reviewers.

There is no difference between not reading and not being able to read.

"Five series purportedly in the multiproxy network (fran003, ital015, ital015x, spai026 and spai047 in the MBH98 list “ITRDB –Miscellaneous”) cannot be located in either the MBH98 collated set or the proxy PC compilations."

Covered in the errata list.

handwaving since that is the most serious discovery and acknowledged by multiple independent statistics experts.

I have the equivalnt of a masters in statistics. But because I have never worked with PCA, I do not know as much about it as I would like. That's one of the things I plan to post. What are your qualifications?

[Guest]

There is a peer review refutation of the MM paper which can be found here;

http://www.realclima...alJClim2004.pdf

Points of note;

1. page 13, 2nd paragraph (through top of page 14)

It should be noted that some falsely reported putative errors in the Mann et al.(1998) proxy data claimed by McIntyre and McKitrick (2003) are an artifact of (a) the use by these latter authors of an incorrect version of the Mann et al. (1998) proxy indicator dataset, and ( their misunderstanding of the methodology used by Mann et al. (1998) to calculate PC series of proxy networks over progressively longer time intervals. In the Mann et al. (1998) implementation, the PCs are computed over different time steps so that the maximum amount of data can be used in the reconstruction.

For example, if a tree-ring network comprises 50 individual chronologies that extend back to AD 1600 and only 10 of those 50 extend to AD 1400 then calculating one set of PCs from 1400 to 1980 (the end of the Mann et al. (1998) calibration period) would require the elimination of 40 of the 50 chronologies available back to AD 1600. By calculating PCs for two different intervals in this example (1400-1980 and 1600-1980) and performing the reconstruction in a stepwise fashion, PCs of all 50 series that extend back to AD 1600 can be used in the reconstruction back to AD 1600 with PCs of the remaining 10 chronologies used to reconstruct the period from 1400-1600. The latter misunderstanding led McIntyre and McKitrick (2003) to eliminate roughly 70% of the proxy data used by Mann et al. (1998) prior to AD 1600, including 77 of the 95 proxy series used by Mann et al. (1998) prior to AD 1500. This elimination of data gave rise to spurious, anomalous warmth during the 15th century in their reconstruction, sharply at odds with virtually all other empirical and model-based estimates of hemispheric temperature trends in past centuries (see e.g. Jones and Mann, 2004).

2. page 32, beginning 2nd sentence, 3rd paragraph:

We compared the RegEM reconstruction with the Mann et al. (1998) surface temperature reconstruction employing the same predictor network, the same calendar annual target season, and same global target region as Mann et al. (1998). We eliminated the infilled values from AD 1400-1403 used by MBH98 to complete one of the Jacoby and D’Arrigo (1989) ‘Northern Treeline’ series back to AD 1400. This is easily done in the RegEM method by treating those values as missing, something that could not have been done in MBH98. We terminated the calibration period in 1971 to address the criticism by McIntyre and McKitrick (2003) of the use by MBH98 of a modest number of infilled missing proxy values in the PC/Multiproxy network between 1971 and 1980.(they still got the same result - Br Cornelius).

3. page 33, beginning 2nd paragraph (key passages highlighted for emphasis):

A remarkably close similarity is observed (Figure 3) between the RegEM and Mann et al. (1998) NH annual mean surface temperature reconstructions. The two reconstructions are indistinguishable well within their 2-sigma uncertainties. The RegEM NH reconstruction using all available individual proxy records (rather than replacing spatially dense tree-ring networks with their leading principal components as in the MBH98 Multiproxy/PC network) again yields nearly indistinguishable estimates (Figure 2). The close reproducibility of the MBH98 reconstruction based on both (a) the use of an independent CFR method and ( the use of the individual proxies used by MBH98 rather than the Multiproxy/PC representation used by MBH98, discredits the arguments put forth by McIntyre and McKitrick (2003) in support of their putative correction to the MBH98 reconstruction.

Figure 1. From Fig. 2, Rutherford et al (2005). Comparisons of stepwise non-hybrid and hybrid-20 reconstructions for annual mean using the multiproxy/PC network and the hybrid-20 reconstruction using the full multiproxy network. Also shown is the instrumental record consisting of the few available data from 1750-1856 (Mann, 2002a) and the Jones et al. (1999) data for 1856-2000.

Figure 2. From Fig. 3, Rutherford et al (2005). Comparison between the RegEM-based hybrid-20 annual mean reconstruction (using the Mann et al. multiproxy/PC network) and Mann et al. (1998) showing overlapping uncertainties in both estimates. Instrumental record is shown for comparison (see Figure 1 for details).

But Little Fish should know this already since we have repeatedly discussed the fact that the Hockey Stick has been confirmed by multiple teams using multiple datasets and different methodologies.

MM went looking for problems but not for the underlying methodology been employed by Mann. If they had attempted to understand the methodology maybe they wouldn't have made so many errors in their critique. I suspect not though since their purpose was to decieve those ill equipped to understand the complexities of statistical analysis. It seems they succeeded despite their errors.

He will also be aware that MM made the bogus claim that the methodology produced false hockey sticks - but they failed to reveal that they went searching for a tiny fraction of the analysis runs which did so, whilst ignoring thousands of examples where no such trend was found - thats right they cherry picked to make a point. Now where have I seen that done before.

The real question to ask is did the small number of errors made in Manns original analysis effect the outcome of that analysis. Rutherford demonstrates - emphatically not. Statistics is designed to cope with a small degree of data errors.

As I suggested Little Fish could waste all our lives re spinning the same denialist bull****.

Br Cornelius

Edited October 25, 2012 by Guest

[Doug1029]

There is a peer review refutation of the MM paper which can be found here;

Br Cornelius

You're saving me some writing.

Little Fish has already decided that anything that disagrees with her preconceived ideas can't be "scientific."

Doug

[Little Fish]

your response is obfuscation.

are you hoping the issue will be lost amongst all your technicals?

here is McIntyre & Mckitrick criticism of Mann:

"(a) unjustified truncation of 3 series"

here is your criticism response to McIntyre & Mckitrick (M&M) point (a):

A series is just the list of ring widths produced in the reading of one tree's rings. M&M do not say what Mann's justification was, nor do they give one of their own. In their unsubstantiated OPINION this was a mistake, but they do not support their claim.

I have truncated series - regretably, I had to truncate my oldest series. The last four rings were pitch loaded and I could not be sure that they weren't false rings. Also, it takes a minimum number of series, at least thirteen, to get a strong enough expressed population signal (EPS must be greater than 0.85.) to be acceptable for climate work. The farther back you go in any chronology, the fewer series there are with which to estimate average ring thickness. Thirteen is the minimum standard, but it usually takes more to get the 0.85 EPS. I have several chronologies with over a hundred years of data that can't be used because I can't guarantee their accuracy due to the small sample size. When part of a series does not intercorrelate with the other series (35% minimum correlation), it is truncated. The good part is used while the rest is discarded."

so in brief you are saying it is ok to truncate a tree ring series because of various issues. I have no problem with that as long as the reasons are valid and documented.

but let's look at M&M are ACTUALLY saying with regards to point (a):

to quote the relevant passage from M&M:

"(a,f) Series #10 and #11 (Central England and Central Europe air temperatures

respectively) use June-July-August averages. This raises three concerns: annual data

were available in the primary sources; other station temperature series used by

MBH98 (#21- #31), where identified, are annual; and MBH98 claims to calculate an

annual temperature index. The Central England Temperature series is truncated at

1730 rather than the available 1659 in source data, which removes a major late 17th

century cold period (see Supplementary Information). Series #10 has a 1987 value

which is 0.43 deg C higher than in the source data though this does not appear to affect

any calculations discussed herein. Central Europe (#11) is truncated at 1550 rather

than the available 1525, which removes the warmest temperatures in the series

(compare Figure 1 Top and Bottom panels). #11, which is an exceptionally long series

of direct temperature information, also shows a notable lack of 20th century

uniqueness. In series #100, MBH98 also crop two very high values from the start of

the series. These truncations are not justified and were not disclosed by MBH98."

you can check the above to be true in the M&M paper here:

http://www.multi-sci...e-mckitrick.pdf

now here is your confused response when I said to you M&M were referring to truncation of instrumental records, not tree rings:

the truncated series referred to are instrumental series, not tree rings chronologies.

how did you make this mistake? LOL.

According to Mann et al. (1998), the series were "multiproxy series" which included "annual resolution dendroclimatic, ice core, ice melt and long historical records combined with other coral, ice and instrumental records." "Dendroclimatic" records are derived from tree rings.

This is a direct quote from Mann's errata sheet to the original Supplementary Information. It was published in Nature:

"Series (2) used in ref. 1 but not listed in original Supplementary Information:

Unpublished Southwest US/Mexico Density series (D. W. Stahle, personal communication)

Unpublished Southwest US/Mexico Latewood Width series (D. W. Stahle, personal communication)"

David W. Stahle is a dendrochronologist at the University of Arkansas and was an ah hoc member of my Ph.D. committee. That Texas-Mexico series is based on TREE RING DATA.

series #10 is Central England Temperature and series #11 is Central Europe Historical. both of these are instrumental records, not tree rings, do you accept the series (10 & 11) referred to by M&M are based on instruments?

your criticism of point (a) is not valid. do you retract it?

[Guest]

Little Fish the justification for truncation was addressed by Doug and in the Rutherford paper.

Truncation is allowed because it permits the inclusion of more data - the other alternative is to chuck that data away - making the overall dataset less robust.

The fact that you are parroting the misunderstanding of MM regarding the need for limited truncation is not really material here.

It is you who are indulging in obfuscation.

I wont quote from it, but here is a detailed analysis of why the MM paper set out to deceive, very clearly written and should enlighten you to what they set out to achieve with each of their critical points;

http://www.realclima...tford-delusion/

Br Cornelius

Edited October 25, 2012 by Guest

[Guest]

The real question is, since MM fail so abysmally to refute the Mann Hockey stick, is there any other compelling refutation out there ??

Br Cornelius

[questionmark]

The real question is, since MM fail so abysmally to refute the Mann Hockey stick, is there any other compelling refutation out there ??

Br Cornelius

The defacto development since then would be good enough, it shows that the hockey stick has developed a kink lately, and the reason is quite evident too: the cold reserves of the North Pole and Greenland are melting and will keep temperature stable until the ice is gone. After that the party continues in the Northern Hemisphere.

[NatureBoff]

Excelllent news. This ties in with the millennial climate cycle of around 1,470 years.